University of Nebraska at OmahaDigitalCommons@UNO

Student Work

2-1991

Otto Rank's Personality Types and the Prisoner'sDilemma GameJeffrey L. CharvatUniversity of Nebraska at Omaha

Follow this and additional works at: https://digitalcommons.unomaha.edu/studentwork

This Thesis is brought to you for free and open access byDigitalCommons@UNO. It has been accepted for inclusion in StudentWork by an authorized administrator of DigitalCommons@UNO. Formore information, please contact unodigitalcommons@unomaha.edu.

Recommended CitationCharvat, Jeffrey L., "Otto Rank's Personality Types and the Prisoner's Dilemma Game" (1991). Student Work. 246.https://digitalcommons.unomaha.edu/studentwork/246

Otto Rank's Personality Types and the Prisoner's Dilemma Game

A Thesis Presented to the

Department of Psychology and the

Faculty of the Graduate College University of Nebraska

In Partial Fulfillment of the Requirements for the Degree

Master of Arts University of Nebraska at Omaha

byJeffrey L. Charvat

February 1991

UMI Number: EP72887

All rights reserved

INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed,

a note will indicate the deletion.

Dissertation Publishing

UMI EP72887

Published by ProQuest LLC (2015). Copyright in the Dissertation held by the Author.

Microform Edition © ProQuest LLC.All rights reserved. This work is protected against

unauthorized copying under Title 17, United States Code

ProQuest LLC.789 East Eisenhower Parkway

P.O. Box 1346 Ann Arbor, Ml 48106- 1346

ThesisAcceptance

Acceptance for the faculty of the Graduate College, University of Nebraska, in partial fulfillment of the requirements for the degree Master of Arts, University of Nebraska at Omaha.

Committee

DepartmentName

Chapman2/26/ 9/Da'te

Table of ContentsIntroduction ................................................... 1M e t h o d ...........................................................11R e s u l t s ..................................................... 2 2D i s c u s s i o n ...................................................... 28R e f e r e n c e s ...................................................... 35Appendix A ...................................................... 38Appendix B ...................................................... 39Appendix C ...................................................... 40Appendix D ...................................................... 41Appendix E ...................................................... 42Appendix F ...................................................... 43Appendix G ...................................................... 44Appendix H ...................................................... 45Appendix I ...................................................... 46Appendix J ...................................................... 47Appendix K ...................................................... 48Appendix L ...................................................... 49

AbstractThis study assesses the validity of Otto Rank's (1929, 1945)personality types and the conceptual systems of 0. J. Harvey (1966). Fifty subjects were selected by faculty members from the Art, Music, Dramatic Arts, and Dance Departments, and the Writers' Workshop, on the basis of personality descriptions from Otto Rank's and 0. J. Harvey's theorizing. Each subject responded to a version of the Prisoner's Dilemma Game against either a cooperative response sequence, a competitive response sequence, or a 50% cooperative/50% competitive response sequence. Subjects completed six measures of cognitive abilities, creativity, and anxiety including the Raven Advanced Progressive Matrices (Set I), Barron-Welsh Art Scale, Nelson-Denny Vocabulary Test, Manifest Anxiety-Defensiveness Scale, Harvey's "This I Believe" Test, and an anagram task.

Statistical analyses were based upon three different groupings of subjects: first, Rank's three personality types (the Artist, the Neurotic, and the Average Person); second, two of Harvey's Conceptual Systems; and third, four groups derived from the creativity ratings provided by the nominating faculty members.

Analysis for Rank's personality types demonstrated a significant increase in competitiveness across trials for both the Neurotic and Average types in the Prisoner's Dilemma Game with Artists showing no change in

competitiveness across trials. No differences were found among Rankian types on the cognitive and personality measures. However, the Artists were rated significantly higher on creativity level than either the Neurotic or Average types.

Significant differences were found between Harvey's System 1 and System 4 subjects in the cooperative and 50/50 conditions of the Prisoner's Dilemma Game with System 1 subjects increasing significantly in competitiveness across trials and System 4 subjects showing no changes in competitiveness across trials. A significant difference was found between these two systems on the Raven Matrices with System 4 subjects performing better than System 1 subjects indicating greater complexity of psychological development among the System 4 subjects. System 4 subjects were also rated significantly higher in creativity than System 1 subj ects.

No differences were found among the four groups based on creativity rating using the Prisoner's Dilemma Game. Significant differences were found among groups on the Nelson-Denny Vocabulary Test with the next to lowest group scoring significantly lower than the other groups. The group rated lowest in creativity used significantly more one-syllable words in their writing than the highest rated creativity group. The meaning of the results are discussed.

1Introduction

The primary aim of this study is to examine the usefulness and validity of Otto Rank’s personality types and to assess their predictive ability using a task designed to differentiate between cooperativeness and competitiveness. Only a few studies have attempted to empirically test the usefulness of Rank's personality types, but the results have all been positive (Helson, 1973a, 1973b; Helson &Crutchfield, 1970a, 1970b; MacKinnon, 1965). In addition,two studies have been performed which, though directed at evaluating Bakan's (1966) theory, have also provided some support for Rank (Brown & Marks, 1969; Carlson, 1971).

Rank (1929, 1945) sees all human behavior as the result of the conflict between two great forces; the fear of life and the fear of death. The fear of life may be understood as the fear of separation, individuality, or loneliness.The trauma of birth is the prototype for this fear as it is an experience of separation from the warmth and security of the womb. Other experiences of this type include weaning from mother's breast and leaving home for the first time.For Rank, the fear of death is the fear of union, fusion, or dependency. One can easily see that these are opposing fears in which some amount of conflict is inevitable. For example, when one considers leaving home for the first time one must choose between the fear of life (separation from familiar people and things) and the fear of death (failing

2to leave and so remaining dependent.) Rank makes it clear that it is one’s conscious choice between these opposing fears that determines which personality type one will manifest.

Rank proposes the existence of three distinct personality types based upon how one balances the conflict between the fear of life and the fear of death. The mechanism that facilitates this compromise is what Rank calls the will or the integrative power of the self. It is first expressed during childhood as counterwill.Counterwill is manifested in young children's attempts to establish separateness from their parents by such events as saying no and rebelling against parents' wishes. It is the different reactions of the parents and the type and amount of support provided to the child that helps or hinders the child's development. The child whose parents accept him or her lovingly and accept the child's expressions of counterwill as normal manifestations of the child's development will grow up to embody the personality type which Rank calls the Creative type or the Artist. The Artist has successfully balanced the fear of life with the fear of death and so expresses his or her individuality while at the same time being capable of integration and union with the rest of society. As such, the Artist represents optimum development.

The second personality type, known as the Conflicted,

3or Neurotic type, expresses the tendency toward separation while denying the need for union because of a greater fear of death than fear of life. This type is the result of the parents' refusal to accept the child's expressions of counterwill. As a result, the Neurotic type will continue to manifest counterwill throughout life and never develop the mature will necessary for successful integration of one's self. The Neurotic will tend to be hostile, negative, critical, arrogant, isolated, and guilty.

The last, and least well-adjusted, personality type is the Adapted type or Average person. The Average person manifests the tendency toward union while denying the tendency toward individuation, thus expressing a stronger fear of life than a fear of death. This person does not demonstrate a strong counterwill as a child but instead finds it easier to maintain union with the parents by simply doing what is expected. For this reason, the child never realizes the possibility of true individuality. This person is characterized by conformity, dependability, superficiality, suggestibility, and lack of dissatisfaction.

Research has provided some support for Rank's three personality types. For example, MacKinnon (1965) used the MMPI, CPI, Myers-Briggs Type Indicator, and several other personality measures to assess a sample of architects who were grouped on the basis of low, moderate, and high creativity. The results showed that the most

4creative architects exhibited personal characteristics consistent with Rank’s Artist type. The group of architects exhibiting moderate creativity corresponded well with the characteristics of the Neurotic type. The least creative group of architects demonstrated personal characteristics consistent with the Average person. Helson and Crutchfield (1970a, 1970b) and Helson (1973a, 1973b) obtained similarresults when examining creative mathematicians and creative writers respectively.

Brown and Marks (1969) designed a study to test Bakan's (1966) constructs of agency and communion. These constructs are similar enough to Rank's as to be useful here. Agency is similar to the fear of death and communion is similar to the fear of life. By using a questionnaire to measure these two tendencies in maladjusted and normal subjects, the investigators found that the maladjusted subjects scored higher on unmitigated agency. This corresponds well with Rank's conception of the Neurotic as one who manifests a strong fear of death and so fails to achieve union.

The goals of the present study are two-fold. First, the work of MacKinnon (1965) was extended using psychological inventories to assess the characteristics of the three Rankian types in three areas of functioning: level of richness and complexity of psychological development, socialization and interpersonal behavior, and evidence for conflict and emotionality. Second, the behavior of Rank's

5three personality types regarding degree of cooperativeness versus competitiveness as measured by the Prisoner's Dilemma Game (Colman, 1982a, 1982b) was assessed. This aspect of Rank's personality theorizing has not been empirically tested until now.

The Prisoner's Dilemma Game (PDG) can be illustrated by the following popular story presented by Luce and Raiffa (1957): Two persons are taken into custody and separated. They are suspected of having committed a serious crime for which the district attorney feels there is insufficient evidence for conviction. A confession is needed from at least one of the suspects for conviction. Each suspect has two options, to confess or not to confess. The district attorney advises each suspect that failure to confess will result in conviction on a lessor charge, such as illegal possession of a weapon, and prison time of one year each.If they both confess to the crime, which the police are confident they have committed, then each will spend eight years in prison. If, however, one suspect confesses and the other does not, then the confessing suspect will receive lenient treatment and spend only six months in jail while the other suspect will get "the book thrown at him" and spend 20 years in prison. Since each must make a decision without knowledge of the other's intended action, neither suspect can be sure whether the other intends to cooperate or to compete for a lighter sentence.

6Adaptation of the PDG for the laboratory can be easily

accomplished by replacing the "payoff” of prison time by a point system. Subjects must decide whether to cooperate to insure mutual accumulation of points or to compete in an attempt to maximize one's own points while minimizing the accumulation of points by the other subject. The "chicken" version of the game is used in this study. It differs from the standard PDG in that the competitive move does not completely dominate the cooperative move.

Dera, Revenstorf, Heyse, and Fitting (1977) found that they could successfully use the PDG to differentiate between subjects classified as socially insecure, hysterical compulsive, anacastic compulsive, and control group subjects. These four groups responded significantly different on five measures of decision-making: risk proneness, decision confidence, decision latency, total risk efficiency, and cooperativeness. Hokanson, Sacco, Blumberg, and Landrum (1980) used a modified PDG to compare depressed individuals, nondepressed individuals with other psychological problems, and normals. They found depressed individuals' interactive patterns to be relatively exploitive and noncooperative.

In a study designed to measure racial prejudice among South African students, Tyson, Schlachter, and Cooper (1988) found the PDG a useful tool for identifying subjects' attitudes toward black and white persons. Ferguson and

7Schmitt (1988) used the PDG similarly and found that subjects' responses revealed stereotypes linked to the gender of the other player. Furnham and Quilley (1988) used the PDG to show that subjects with a high Protestant work ethic belief are more competitive than those with a low Protestant work ethic belief, as measured by the Protestant work ethic scale.

These research efforts demonstrate the usefulness of the PDG for investigating individual differences in attitudes and traits of participating subjects. In the present study, subjects were led to believe that they were playing the PDG with another volunteer subject or disinterested observer, but were in actuality playing with one of three levels of a preprogrammed, non-contingent response sequence. One condition involved 75% cooperative responses, the second condition involved 75% competitive responses, and the third condition involved a 50% cooperative and 50% competitive response sequence. Each personality type responded in each of the three PDG conditions.

The PDG was used in the present study to measure cooperative versus competitive behavior between individuals. The PDG is believed to reflect what MacKinnon (1965) called socialization and interpersonal behavior. It was predicted that Rank's three personality types would respond differentially to the PDG. Because the Average person is

8believed to exhibit conformity and adaptation, along with the shunning of individuality it was expected that the Average person would tend to cooperate in all three PDG conditions while the tendency of the Neurotic type to be hostile, negative, and isolated would promote competitiveness in all three conditions.

The Artist's ability to balance differentiation from others and integration with others suggests that the Artistwill act without rigid adherence to conformity or isolationbut to act instead in a flexible manner. It was expected that the Artist would tend to cooperate in the condition in which cooperation is salient and to compete in the condition in which competition is salient. For the Artist, the cooperativeness or competitiveness of the condition is the primary basis for making choices since the Artist is not expected to respond in a stereotyped manner.

Identification of subjects as to personality type was accomplished by providing brief personality descriptions to professors from which they nominated possible subjects. The personality descriptions presented to the nominators were composites of Rank's and Harvey's types. They were adapted from the works of Maddi (1989) and Harvey (1966) . Harvey's Conceptual Systems theory results in four systems ranging from low differentiation, low integration, and highly concrete functioning among System 1 types to high differentiation, high integration, and abstract

9functioning among System 4 types.

For Harvey, System 1 persons believe that truth and reality are external and independent of the observer and that everything is controlled by some supra-personal force (e.g., God, luck, or fate) which determines a person's behavior regardless of his efforts. System 1 persons are poorly differentiated and integrated and rely heavily upon tradition, normative standards, authority figures, and societal laws as their guidelines for action. They tend to exhibit a high degree of fundamentalism of religious beliefs, high ethnocentrism, and high evaluativeness. Their self-worth is determined by the extent of conformity to rules and regulations. These persons are most like Rank's Average type.

System 2 persons are similar to Rank's Neurotic type. They are hostile toward and suspicious of institutional authority and are best characterized by what they argue against. They are opposed to conventionality and emphasize nonconformity. They exhibit considerable cynicism and alienation. Fear and anxiety associated with new experiences usually leads to withdrawal and defensiveness.

System 3 persons are less evaluative than System 1 and System 2 persons but their views tend to be superficial and shallow. They rarely express a strong commitment to a particular course of action or belief. They are concerned mainly with personal acceptance and approval by others,

10especially those of high status and expertise. This need results in a readiness to compromise personal judgment in favor of conformity. They tend to express sentiments that everyone's views are right, that everyone needs everyone else, and that social relationships help one grow and understand oneself.

System 4 persons are the most differentiated and most integrated of the four systems and are similar to Rank's Artist type. They are characterized by a balanced need for both mutuality and autonomy and are the most tolerant of stress and of diverse ideologies and behavior. Their behavior is characterized by high task orientation, information seeking, exploration, independence without negativism, and creativity.

In addition to the PDG, subjects completed the following pencil-and-paper measures of what MacKinnon (1965) called the level of richness or complexity of psychological development: a nonverbal intelligence test (the Raven Advanced Progressive Matrices Set I, Raven, 1965), a measure of concreteness-abstractness (the "This I Believe" Test, Harvey, 1966), an index of one-syllable words usage, a measure of creativity (the Barron-Welsh Art Scale, Barron & Welsh, 1952), a measure of verbal fluency (an anagram task, Mullins, 1978), and a measure of vocabulary (the Nelson-Denny Vocabulary Test, Nelson & Denny, 1960).Subjects also completed a measure of anxiety proneness (the

11Manifest Anxiety-Defensiveness Scale, Millimet, 1970) believed to measure the third category of functioning considered by MacKinnon, that of degree of personal soundness or psychological health.

MethodSubjects and Procedure

Subjects were selected from five departments at the University of Nebraska at Omaha by professors from those departments. The five departments were Art, Music, Dramatic Arts, Dance, and the Writers' Workshop. These five departments were chosen because students in these areas of endeavor are pursuing studies which allow them unique opportunities for creativity and provides their professors a unique opportunity to judge the student's creative efforts as well as how each student's work expresses his or her personality. Faculty in these programs were asked to select only from among those students he or she knew both personally and professionally. Nominators were assured anonymity and were asked not to discuss the selection process with anyone inside or outside their departments.

A total of 30 professors were contacted by mail requesting nominations based upon three personality descriptions provided to the nominators. The descriptions reproduced below are based upon Rank's (1929, 1945) writingsand are adapted from the work of Maddi (1989) and the work of Harvey (1966) whose conceptual systems match remarkably

12well with those of Rank.

The Artist Type These persons exhibit considerable complexity resulting from a high degree of differentiation and integration of thoughts, feelings, and actions.They exhibit intimacy with, and commitment to, other people without slavish loyalties and undue concern for social proprieties. These persons neither control others, nor are controlled by others. They tend to express individuality while at the same time being capable of union and integration with the rest of society. In their work, they are productive in the direction of unusualness but also usefulness. Their behavior is characterized by high task orientation, information seeking, exploration, independence without negativism, and creativity. They are capable of recognizing problems that require attention and solution. They will present several courses of action that may rectify the problem and are not adverse to recommending one of the alternatives as likely to be the most useful.

The Neurotic Type These persons have committed themselves to the pain of separation from the herd but have not developed a constructive integration with the world. Instead

13of expressing mature will or a strong sense of self, they seem fixated at a level of contrariness. They act either against people or completely separately from them. They are rebellious, suspicious, cynical, alienated, and opposed to conventionality. Although their personalities may show much differentiation and their actions may appear guite novel, even clever at times, close examination reveals these persons to possess high stereotypy and an inability to try alternate approaches to complex problems.Their sense of separateness is ridden by hostility, arrogance, isolation, guilt, and undue criticalness.

The Average Type These persons are conforming, dependable, suggestible, superficial, and self-satisfied. They never seriously entertain the possibility of their own individuality. Their truths are illusory and vanish overnight if the social milieu to which they belong shifts its values. They are tolerant, nonjudgmental, and rarely express a strong commitment to any particular course of action or belief. Their central concern is for personal acceptance and approval by others, especially by persons of high status and expertise. In so doing, these persons seek maximum adaptation to, rather than personal transcendence of, the social environment.They espouse the notion that everyone's beliefs are

14right, that everyone needs everyone else, and that social relationships help one grow and understand oneself.

These descriptions were presented to the nominators without being labeled as Artist, Neurotic, and Average types, but as Type 1, Type 2, and Type 3 respectively. Professors who failed to respond within one week were contacted a second time by mail. Ten professors (two from the Art Department, three from the Music Department, two from the Dramatic Arts Department, one from the Dance Department, and two from the Writers' Workshop) responded with a total of 93 nominations, including 18 from the Art Department, 3 8 from the Music Department, 19 from the Dramatic Arts Department, 4 from Dance, and 19 from the Writers' Workshop. A total of 41 Artists, 26 Neurotics, and 27 Average types were identified for participation. Agreement between at least two raters as to personality type was initially planned as a requirement for including a nominee in the study, however, only a minority of persons were nominated by more than one professor so this requirement had to be dropped. Three subjects who were nominated by more than one faculty member to different Rankian categories were eliminated from further consideration.

Nominees were contacted by telephone and informed that they had been nominated by a professor from within their respective departments to participate in a study of

15creativity. Numerous attempts were made to contact each subject over a period of three months. Of the initial set of 93 nominees, 26 could not be contacted, 12 agreed to participate but were unable to attend, five refused to participate, and 50 agreed to participate of which 19 were nominated as Artists, 15 as Neurotics, and 16 as Average types. Of the 50 subjects participating, 28 females and 22 males, 10 were nominated by more than one professor to the same personality type. Subjects ranged in age from 18 to 55 (X — 2 6.7, SD=7.51) and represented educational levels ranging from freshman to Ph.D. All subjects volunteered their time and effort. See Appendix A for a breakdown of subjects by department, Rankian personality type, conceptual system, and creativity rating. Information regarding the selection process was kept strictly confidential.

Following informed consent, subjects completed seven pencil-and-paper tests either in groups or individually, and also participated individually in the Prisoner's Dilemma Game. The tests are presented below in the order of their administration.Raven Advanced Progressive Matrices (Set I)

This test consists of 12 large geometric patterns with each missing one small part of the design. Following each pattern are eight small designs, only one of which appropriately completes the large pattern. Subjects were to pick the design which best completes the pattern. They were

16given 10 minutes to complete this test. This test is used as a quick indicator of intelligence (Raven, 1965).The "This I Believe" Test

The "This I Believe" Test (TIB) is based upon the work of Harvey (1966) and was used to identify subjects according to Harvey's four levels of conceptual system functioning. Each subject was asked to write as much as he or she could on nine topics presented in the form of the statement "ThisI believe about _________ ." Subjects had two minutes towrite on the first few topics with the time allotted reduced with each topic so that by the end of the test each subject was allowed one minute and forty-five seconds for each response. The reason for the time reduction is to insure that subjects will not have time to prepare socially desirable responses and will provide responses that are in line with their true beliefs. The test consists of nine subjects: the American way of life, religion, people,marriage, friendship, sin, revenge, lying, and calling ateacher by his or her first name. Greaves (1971) presented evidence that the test is reliable and stable.One-svllable Words

This is a measure of writing fluency (Mullins, 1977). The score is determined by counting the number ofone-syllable words in the first 100 words each subject wroteon the TIB.

17Barron-Welsh Art Scale

This scale is a portion of the Welsh Figure Preference Test (Barron & Welsh, 1952). It consists of 86 black and white drawings. Subjects are asked to work as fast as theycan on deciding whether or not they like each drawing. Ifthey are unsure of any drawing, they are instructed toguess. No time limit is given. This test was used byMacKinnon (1965) as a measure of creativity.Nelson-Dennv Vocabulary Test

This measure of verbal ability is a part of the Nelson-Denny Reading Test (Nelson & Denny, 1960). Subjects were asked to choose the answer among five alternatives which best completes the stimulus phrase (e.g., A dog is ... 1. a reptile 2. a plant 3. a stone 4. an animal 5. a book). Subjects are given ten minutes to complete the 100 items composing the test."Generation" Anagram Task

This task required subjects to compose as many four- or more-letter words as possible in five minutes from the word "generation" using each letter only once. This task has been used as a measure of word fluency and spatial ability (Mullins, 1977).Manifest Anxietv-Defensiveness Scale (MAD)

This measure of anxiety proneness consists of a male and female version composed of 63 and 59 items respectively (Millimet, 1970). Both versions of this scale correlate

18extremely high (r=.92) with the Taylor Manifest Anxiety Scale, a scale used by MacKinnon (1965). The MAD scale has been shown to possess very high reliability and satisfactory validity (Millimet, 1970). An illustrative item is "Most nights I go to sleep without thoughts or ideas bothering me." Subjects are asked to read each statement and decide whether it is true as applied to them or false as applied to them.Prisoner's Dilemma Game (Chicken Version)

Each subject participated in the Prisoner's Dilemma Game against one of three preprogrammed, non-contingent response sequences of three blocks of 16 trials each. The first condition was composed of 75% cooperative responses. The second condition was composed of 75% competitive responses. The third condition was composed of 50% cooperative responses and 50% competitive responses.Subjects composing each of Rank's three personality types were fully crossed with the three conditions of the PDG.

Upon arrival, subjects were shown two side-by-side rooms, each containing a Zenith microcomputer terminal, a table, and a chair. The terminals appeared to be connected by wires through a small opening in the wall between the rooms. The PDG program was on a floppy disc with each program containing each of the three experimental conditions. Both computers were programmed to play the same response sequence and subjects were encouraged to choose one

19of the rooms for their participation. Subjects were led to believe that they were to play the game with another subject or a disinterested bystander. They were informed that the other player was waiting nearby and that they would position themselves in the other room while the subject was reading the instructions and becoming familiar with the payoff matrix. Subjects were informed that they were not to see the other player in order to avoid biasing their performance. They were told, for example, that their knowledge as to the sex of the other player could have an unwanted effect upon their performance. Subjects were allowed to choose either room in order to foster the illusion of interaction with another player.

A standardized explanation of the game was given both in written form and then orally to insure that each subject understood how the game was to be played and how the points were to be allocated. A payoff matrix was included with the written instructions. Subjects were to respond on each trial with a choice of either "A" or "B." Subjects did not respond to a question or statement but were to choose a strategy based entirely upon the payoff matrix and responses of the other player. The payoff matrix indicated that a choice of "A" by both players would lead to a payoff of five points for each player. A choice of "B" by both players would lead to a loss of four points for each player. Should one player choose "A" and the other choose "B, 11 the player

20choosing "A" would lose three points and the player choosing "B" would gain six points. After each trial, subjects were informed of the other player's choice, the point allocations for each player on that trial and a cumulative point total for each player. After the sixteenth trial, subjects were informed that the scores would be reset to zero and point accumulation would begin again. Appendix B is a reproduction of the written instructions and payoff matrix. Verbal instruction involved repetition of the written instructions.

Subjects were informed that neither player would have advance knowledge of the other's choices. They were told that they were to play the game in whatever way they felt was most appropriate and leading words such as "opponent," "cooperate," and "compete" were avoided. Subjects were referred to as players or co-players. Each subject was led to believe that another subject or disinterested bystander was playing the "other side."

Subjects were assigned to one of the three conditions in a systematic manner (by type and sex) to insure that each personality type would be represented in each condition. Subjects were matched according to gender in order to eliminate any effects of gender differences in performance.

Once a subject was positioned at the terminal in one of the rooms and had begun to read the instructions, the experimenter excused himself for the supposed purpose of

21helping the other player to get set up. The experimenter entered the other room and made enough noise to convince the subject that another player was indeed in the other room.The experimenter returned to the subject's room to answer any questions the subject may have had. Care was taken to insure that information as to the purpose of the game was not revealed so that each subject had to decide for him- or herself the purpose of the game and the strategy to be employed. The subject was told to begin by the experimenter who assured the subject that he would be outside the door if any problems arose.

Appendices C and D contain further information regarding the PDG program such as order of preprogrammed responses, verbal prompts to the subject to make a choice, and feedback information. After completion of the game, subjects were probed to see if they had suspected the absence of another player in the other room. Subjects were debriefed as to the purpose of the study, how well they had performed, and thanked for their participation.

Following completion of the PDG and administration of the pencil-and-paper measures (approximately four months after the original nomination procedure), each subject was rated by the same faculty member as to creativity level based upon a scale of 1 (low talent) to 11 (high talent). Appendix E presents the form used.

22Results

Analyses were performed based upon three different groupings of subjects. First, comparisons were made using Rank's personality types based upon nominations using the personality descriptions. Second, comparisons were made on the basis of conceptual system functioning as identified by the responses of subjects on the TIB. And third, four groups were compared based upon the ratings of creativity supplied by the original nominator. Each of the three groupings of subjects was compared on the PDG using multivariate analyses of variance and on the personality measures using one-way analyses of variance. One-way analyses of variance were also used to analyze creativity ratings in regard to Rank's personality types and Harvey's conceptual system types.

Analysis of the TIB was accomplished with two raters independently reading each TIB and arriving at a decision as to which of the four categories of Harvey's Conceptual System approach was appropriate. A 60% agreement rate for the TIB was obtained, after which both raters worked together to arrive at a consensus for each subject with respect to conceptual system functioning. System 2 and System 3 subjects were not included in all remaining analyses as too few were identified by the TIB.

A chi-square analysis between Rankian type and Conceptual System showed that the Average type was most

23likely to be identified as System 1 (12 out of 15), while the Artist type was likely to be identified as System 4 (11 out of 16), X2 (l)=8.58, p<.01.Analyses for Rank’s Three Personality Types

A 3 (personality type) X 3 (PDG condition) X 3 (PDG trial blocks) factorial analysis of variance with repeated measures on the last factor was used to analyze the number of competitive responses made by subjects in each of the three PDG conditions and across the three trial blocks of 16 trials each. Analyses of the pencil-and-paper measures were accomplished by using one-way analyses of variance and the Tukey A multiple comparison procedure.

Neither the main effects of Personality Type,F (2,41) =.05, p = .96, nor PDG condition, F(2,41)=.67, p=.52, were statistically significant. However, the main effect for Trials, F (2,82)=13.05, p<.001, and Personality Type X Trials interaction, F (4,82)=2.60, p=.042 were statistically significant. The analysis of variance table is presented in Appendix J. A simple effects analysis of the interaction revealed a significant increase in competitiveness for the Neurotic type across trials, F (2,82)=7.74, p<.01, and for the Average type across trials, F (2,82)=7.15, p<.01. Tukey A multiple comparisons analysis of these effects revealed an increase in competitiveness for the Neurotic type between trial block 1 (X=7.80) and trial block 3 (X=10.60), p<.01,and an increase in competitiveness for the Average type

24between trial block 1 (X=7.96) and trial block 2 (X=10.27),p < . 01, and between trial block 1 and trial block 3 (X=10.16), p < .01. No significant changes were seen in competitiveness for Artists across the three trial blocks.

The expected three factor interaction of Personality- Type X PDG condition X PDG trials was not significant,F ( 8,82)=1.16, p = .336.

No significant differences were found for Rank's personality types on the remaining pencil-and-paper measures: Raven Advanced Progressive Matrices, F(2,47)=.15, p=.86; Nelson-Denny Vocabulary Test, F (2,47)=2.19, p=.12; anagram task, F(2,47)=.30, p=.75; Barron-Welsh Art Scale, F(2,47)=.27 p=.76; Anxiety Proneness, F(2,47)=.41, p=.67; one-syllable words, F(2,47)=.34, p=.71.

A one-way analysis of variance on the creativity ratings was statistically significant, F (2,47)=5.57, p = .0067. A Tukey A multiple comparisons analysis revealed that the Artist type was rated significantly more creative than either the Neurotic (p<.05) or Average (p<.05) types. The mean creativity rating for Artists was 8.53, for Neurotics 6.47, and for Average types 6.13. No significant difference in creativity level was found between the Neurotic and Average types. Appendices F and G present summaries of these data.Analyses for Harvey's Conceptual Systems

Analyses based upon Harvey's systems as identified by

25the TIB were completed in a similar manner. System 2 (n=6)and System 3 (n=2) subjects were not included in theanalysis as an insufficient number were identified by the TIB.

The number of competitive responses in the PDG were compared using a 2 (Conceptual System) X 3 (PDG condition)X 3 (PDG trial blocks) factorial analysis of variance with repeated measures on the last factor. The analysis of variance table may be found in Appendix K. Neither the main effect of Conceptual Systems, F(l,36)=.02, p=.89, nor Condition, F(2,36)=.30, p=.75, were statistically significant. The main effect for Trial Blocks,F (2 ,72)— 1 .08, p = .0 02 , Conceptual Systems X Trial Blocks interaction, F (4,72)=2.95, p=.059, and the Conceptual Systems X PDG Condition X Trial Blocks interaction,F (4,72)=2.80, p = .032, were statistically significant.

Analysis of simple effects of the three factor interaction revealed a significant effect for System 1 subjects in the 75% cooperative condition across trial blocks, F (2,72)=8.71, pc.01. The Tukey A procedure revealed a significant increase in competitiveness between trial block 1 (X=7.80) and trial block 2 (X=10.60), pc.01, and between trial block 1 and trial block 3 (X=11.00), pc.01.The difference between trial block 2 and trial block 3 was not statistically significant.

There was also a significant effect for System 1

26subjects across trials in the 50/50 condition, F (2,72)=6.36, p < .01. The Tukey A procedure revealed a significant increase in competitiveness between trial block 1 (X=7.60) and trial block 3 (X=11.80), pc.01. In the 75% cooperativecondition, System 1 subjects (X=11.417) were significantly more competitive than System 4 subjects (X=7.333) on the third trial block, F (1,63)=4.27, pc.05.

A significant difference was found between System 1 subjects and System 4 subjects on the Raven Advanced Progressive Matrices, F (1,40)=4.44, pc.05, with System 1 subjects having a mean of 9.96 and System 4 subjects a mean of 10.84. A significant difference was also found between the creativity ratings for the System 1 subjects (X=6.43) and System 4 subjects (X=8.05), F (1,40)=5.03, p=.0305.

The one-way analyses of variance for the Nelson-Denny Vocabulary Test, F(l,40)=.14, Barron-Welsh Art Scale,F (1,40)=1.97, anagram task, F (1,40)=1.60, Manifest Anxiety-Defensiveness Scale, F (1,40)=1.77, and one-syllable words procedure, F(l,40)=.19, were not statistically significant. See Appendices F and H for the group means. Analyses for Four Groups Based Upon Creativity Ratings

The creativity ratings supplied by the faculty were used to generate four groups of subjects and were analyzed in the same manner as Rank's types and Harvey's Conceptual Systems on the PDG and the paper-and-pencil measures. Group 1 (n=14) consisted of subjects with ratings of 2 to 5; Group

272 (n=12), consisted of subjects with ratings of 6 or 7;Group 3 (n=13), consisted of subjects with ratingsof 8 or 9; Group 4 (n=ll), consisted of subjects with ratings of 10 or 11.

For the PDG, the number of competitive responses were compared using a 4 (Creativity Groups) X 3 (PDG condition) X3 (PDG trial blocks) factorial analysis of variance with repeated measures on the last factor. The main effects for Creativity Group, F (3,38)=1.64, p=.20, and PDG Condition, F(2,38) = .27, p = .77 were not statistically significant. The main effect for Trial Blocks, F (2,76)=7.25, p=.001, was statistically significant. The analysis of variance table may be found in Appendix L. The Tukey A procedure revealed a significant increase in competitiveness between trial block 1 (X=8.20) and trial block 2 (X=9.88), pc.01, andbetween trial block 1 and trial block 3 (X=9.88), pc.01. Nosignificant interaction effects were found.

A one-way analysis of variance of the scores on the Nelson-Denny Vocabulary Test was statistically significant,F (3,46)=5.50, p = .002 6. The Tukey A procedure revealed significantly lower scores for Group 2 (X=35.33) than forall other groups (Group 1, X=4 9.79; Group 3, X=51.38; Group 4, X=4 7.91) .

On the one-syllable words task, Group 1 subjects (X=70.50) used significantly more one-syllable words in the first 100 words of their TIB responses than were used by

28Group 4 subjects (X=62.00), F (3,46)=3.61, p=.0201. The one-way analyses of variance for the Raven Advanced Progressive Matrices, F (3,46)=1.20, Nelson-Denny Vocabulary Test, F (3,46)=5.50, anagram task, F (3,46)=1.14, Barron-Welsh Art Scale, F(3,46)=.87, Manifest Anxiety-Defensiveness Scale, F (3,46)=1.19, were not statistically significant.See Appendices F and I for a complete summary of these data.

DiscussionRank's three personality types were expected to

perform differently from each other on the PDG. The Average type was expected to cooperate in all three conditions, theNeurotic type was expected to compete in all threeconditions, and the Artist type was expected to cooperate in the cooperative condition and compete in the competitive condition. These expectations were not supported by the analyses. In fact, both the Neurotic-type subjects and the Average-type subjects increased in competitiveness across trials. No increase in competitiveness was seen for the Artist.

Analyses of the personality measures were expected to reveal that the Artist possesses greater complexity of psychological development and greater creativity than the Neurotic type and the Average person type. Furthermore, theNeurotic type was expected to score higher on anxietyproneness than either of the other two types. Although the level of creativity as measured by faculty ratings showed

29that the Artist type was rated significantly higher in creativity than both the Neurotic and Average types, none of the other expectations were supported.

Some factors that could account for the lack of significant differences among personality types should be discussed. The selection process may have failed to reliably identify Rank's personality types. In one college department, only one professor agreed to help make nominations. Where more than one professor made nominations in the same department, dual nominations were uncommon. The result was that subjects were allowed to participate in the study without the planned requirement that each subject be nominated to the same personality category by at least two nominators.

In the hope of achieving agreement between two raters, consideration was given to submitting the initial nominees to other faculty members using a forced choice format. Such a procedure was rejected out of the belief that a careful nomination made by a single faculty member who knows a subject very well would produce more reliable results. Nevertheless, 10 of the 50 participants in the study were nominated by at least two faculty members.

Indeed, comparison of Rank's personality types identified via the nomination procedure and Harvey's Conceptual System types identified by an analysis of the "This I Believe" Test resulted in a significant

30correspondence between the two approaches and provides some evidence for the validity of the nomination process. As expected, subjects identified as Average types by faculty nominators were identified as System 1 persons by the TIB analysis, while subjects identified as the Artist type by faculty nominators were identified as System 4 persons by the TIB analysis. Unfortunately, an insufficient number of System 2 persons were identified by the TIB to allow comparison with the Neurotic type.

Analysis of the results based upon Harvey's Conceptual System level of functioning provided support for the hypothesis that the distinctions based upon this system are valid. Although too few System 2 and System 3 subjects prevented these groups from being included in the analyses, significant differences were noted between System 1 and System 4 subjects. The Conceptual System X PDG Condition X PDG Trial Blocks interaction showed that System 1 subjects became increasingly more competitive across trials in the cooperative and 50/50 conditions, whereas the System 4 subjects showed no significant changes in competitiveness in either condition. Furthermore, System 1 subjects were significantly more competitive than System 4 subjects in the third trial block of the cooperative condition.

Why did the System 1 subjects become significantly more competitive across trials in the cooperative and 50/50 conditions and why were they more competitive than the

31System 4 subjects in trial block 3 of the cooperative condition? Consideration of the characteristics of the two systems may provide answers to these questions. System 1 persons are characterized by concrete thinking and high evaluativeness such as good versus bad, right versus wrong, etc. The concrete thought processes of System 1 subjects may have resulted in their lack of appreciation for that aspect of the PDG which allows for a cooperative accumulation of points by both players. System 1 subjects may have concluded that the purpose of the game was to win by earning the most points while giving up the fewest points.

System 4 persons, on the other hand, are believed tothink more abstractly, are expected to be more aware ofsituational and personal causes of events, to be pluralistic in their search for explanations, and therefore to be more aware of the possibility of mutual accumulation of points by cooperating instead of competing. Additional evidence for System 4 persons as more psychologically complex than System 1 persons is seen in the results of the Raven Advanced Progressive Matrices, a nonverbal measure of intellectual ability, in which System 4 subjects performed significantlybetter than the System 1 subjects.

System 4 subjects were also rated significantly higher in creativity than were System 1 subjects. This result is consistent with the outcome of the creativity ratings of the

32three Rankian types, but is perhaps more believable as the assignment to systems was based on the responses of subjects to the TIB and evaluation was performed by persons who were not associated with the initial nomination procedure.

Analysis for groups based upon creativity ratings showed no significant differences on the PDG with the exception of an effect for trial blocks, i.e., subjects in general became more competitive over trials. As noted earlier, four groups were created based upon the creativity rating provided by the faculty nominators. Group 2, the next to lowest creativity group, scored significantly lower on the Nelson-Denny Vocabulary Test when compared with each other group. Subjects in Group 1, the lowest creativity group, used significantly more one-syllable words than Group 4 in responses to the TIB.

Thus, subjects rated lower in creativity exhibited poorer verbal and writing abilities than subjects rated higher in creativity. This is in line with the theoretical positions of both Rank's and Harvey's systems. Both predict that increases in creativity are correlated with greater differentiation and integration in the psychological and social realms. No explanation can be given at this point as to why it was Group 2 and not Group 1 which scored significantly lower than the other groups on the vocabulary test.

Some reasons why the PDG may not have differentiated

33between personality types as well as had been predicted should be discussed. Because the preprogrammed responses against which each subject played followed a fixed pattern, instead of being truly random, some subjects may have suspected that they were not actually playing against another player or that the other player was instructed to play in a certain manner. During the debriefing, the experimenter questioned each subject for such suspicions and it is clear that the majority of the subjects believed that they were in a game with a real person playing the other side. Of the 50 subjects in the experiment, eight expressed some suspicion, though only one expressed outright disbelief. These subjects indicated that they only became suspicious towards the end of the three trial blocks. Nevertheless, future research may wish to incorporate a truly random sequence of responses to insure total believability.

With regard to Rank's types, it may be that the PDG is not a strong enough situation to elicit typical reactions from each of the personality types. For example, although the Average type is expected to cooperate with other people, this may be true only in life situations which are personally meaningful to the person.

Lack of significant findings in the competitive situation may have resulted from the fact that the competitive condition necessarily compels a person to

34compete. Indeed, there is no strategy one can employ in the competitive condition that would result in an outright win. All a competitive response can do against a competitive strategy is minimize one's losses.

The results of this study provide some evidence in support of both Rank's and Harvey's theories. The conception of a dimension of personality based upon a continuum from concrete to abstract thinking and undifferentiation to differentiation and integration appears valid. The "This I Believe" Test and the Prisoner's Dilemma Game have both demonstrated some utility in identifying a person's characteristics with respect to this continuum. These findings have strong practical implications in the identification of those individuals suitable for leadership roles requiring flexibility and creativity.

35References

Bakan, D. (1966). The duality of human existence. Chicago: Rand McNally.

Barron, F., & Welsh, G. S. (1952). Artistic perception as apossible factor in personality styles: Its measurementby a figure preference test. Journal of Psychology. 33., 199-203.

Brown, D., & Marks, P. A. (1969). Bakan1s bi-constructs: Agency and communion. The Psychological Records. 19. 465-478.

Carlson, R. (1971). Sex differences in ego functioning: Exploratory studies of agency and communion. Journal of Consulting and Clinical Psychology. 37., 267-277.

Colman, A. (1982a). Game theory and experimental games:The study of strategic interaction. Oxford: Pergamon.

Colman, A. (1982b). Cooperation and competition in humans and animals. London: Van Nostrand Reinholt.

Dera, M . , Revenstorf, D., Heyse, H . , Fitting, U. (1977).An experimental game as a diagnostic aid. Zeitschrift fur Klinische Psychologies 6, 161-176. (From Psychological Abstracts. 1978, .63, Abstract No. 43)

Ferguson, E. D., & Schmitt, S. (1988). Gender-linked stereotypes and motivation affect performance in the prisoner's dilemma game. Perceptual and Motor Skills,66, 703-714.

Furnham, A . , & Quilley, R. (1989). The Protestant workethic and the prisoner's dilemma game. British Journal of Social Psychology, 28, 79-87.

Greaves, G. (1971). Harvey's "This I Believe" Test: Studies of reliability. Psychological Reports. 2 8 . 387-390.

Harvey, 0. J. (1966). Experience, structure and adaptability. New York: Springer Publishing Co.

Helson, R. (1973a). The heroic, the comic, and the tender: Patterns of literary fantasy and their authors. Journal of Personality, 4 1 , 163-184.

Helson, R. (1973b). Heroic and tender modes in womenauthors of fantasy. Journal of Personality. 4 1 . 493-512.

Helson, R . , & Crutchfield, R. S. (1970a). Creative types in mathematics. Journal of Personality, 38. 177-197.

Helson, R . , & Crutchfield, R. S. (1970b). The creative researcher and the average PhD. Journal of Consulting and Clinical Psychology. 2, 237-250.

Hokanson, J. E., Sacco, W. P., Blumberg, S. R . , & Landrum,G. C. (1980). Interpersonal behavior of depressive individuals in a mixed-motive game. Journal of Abnormal Psychology, 89, 320-332.

Luce, R. D., & Raiffa, H. (1957). Games and decisions. New York: Wiley.

MacKinnon, D. W. (1965). Personality and the realization of creative potential. American Psychologist. 20, 273-281.

Maddi, S. R. (1989). Personality theories: A comparative analysis (5th ed.). Chicago: The Dorsey Press.

Millimet, C. R. (1970). Manifest anxiety-defensivenessscale: First factor of the MMPI revisited. Psychological Reports. 2 7 . 603-616.

Mullins, M. R. (1977). The relationship of verbal abilities to cognitive complexity. Unpublished master's thesis, University of Nebraska at Omaha.

Nelson, M. J., & Denny, E. C. (1960). The Nelson-DennyReading Test, manual

Rank, O. (1929). The trauma of birth. New York: Harcourt, Brace.

Rank, O. (1945). Will therapy and truth and reality. New York: Knopf.

Raven, J. C. (1965). Advanced Progressive Matrices: Sets I and II, manual

Tyson, G. A., Schlachter, A., & Cooper, S. (1988). Game playing strategy as an indicator of racial prejudice among South African students. The Journal of Social Psychology. 128, 473-485.

Appendix ATypes, Systems, and Groups by Department.

38

Art Music Drama Dance Writers1Workshop

RankArtist 4 5 4 2 4Neurotic 4 7 3 0 1Average 0 8 3 0 5Total 8 20 10 2 10

HarveySystem 1 3 15 4 0 3System 2 1 0 3 0 2System 3 1 0 1 0 0System 4 4 5 3 2 5Total 8 20 10 2 10

CreativityOne 4 3 4 0 3Two 3 5 2 0 1Three 1 8 1 1 3Four 0 4 3 1 3Total 8 20 10 2 10

39Appendix BInstructions and Payoff Matrix for the Prisoner's Dilemma Game.

This game has two players and is played over a number of trials,

A trial consists of each player selecting eitherRESPONSE "A" or RESPONSE "B"

Neither player will be informed of the selection of theother player until BOTH players have entered their responsefor that trial.

After both players have made their selection, the responses of each player will be shown on the computer screen of each player at the same time along with points gained or lost by both players.

Points will be awarded to each player on the basis of the following allocation:

YOU: RESPONSE "A11 OTHER: RESPONSE "A11YOU GAIN: +5 OTHER GAINS: + 5

YOU: RESPONSE "A" OTHER: RESPONSE "B"YOU LOSE: -3 OTHER GAINS: + 6

YOU: RESPONSE "B" OTHER: RESPONSE "A"YOU GAIN: +6 OTHER LOSES: -3

YOU: RESPONSE "B" OTHER: RESPONSE "B"YOU LOSE: -4 OTHER LOSES: -4

40Appendix CPrisoner’s Dilemma Game Program Information. Preprogrammed Response Sequences.

Condition A: 75% Cooperative Response Sequence Trial 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16Response A A B A A A B A A A B A A B A A

Condition B: 75% Competitive Response Sequence Trial 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16Response B B A B B B A B B B A B B A B B

Condition C; 50/50 Competitive/Cooperative Response Sequence Trial 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16Response A B A A B B B A B A A A B B A B

41Appendix DPrisoner's Dilemma Game Program Information.

Sample Prompt Before Each TrialYOU ARE PLAYER 2 AND THIS IS TRIAL 1.

PLEASE MAKE YOUR SELECTION: A (TYPE A) OR B (TYPE B)? Sample Feedback Received After Each Trial

THE RESPONSE OF PLAYER 1 IS: A THE RESPONSE OF PLAYER 2 IS: A

PLAYER 1 EARNED 5 POINTS ON TRIAL 1 PLAYER 2 EARNED 5 POINTS ON TRIAL 1

AFTER ALL TRIALS PLAYER 1 HAS EARNED 5 POINTS AFTER ALL TRIALS PLAYER 2 HAS EARNED 5 POINTS

WHEN YOU ARE READY FOR THE NEXT TRIAL: PRESS F5Sample Prompt Between Trial Blocks

THE POINTS ON THE PRECEDING TRIALS WILL BE SET TO ZERO FOR BOTH PLAYERS AND THE GAME WILL BEGIN ANEW.WHEN YOU ARE READY FOR THE NEXT TRIAL: PRESS F5

Sample Prompt at the End of the GameTHE GAME IS OVER. DO NOT STRIKE ANY KEYS.

PLEASE STAY SEATED AND WAIT FOR THE EXPERIMENTER. THANK YOU FOR YOUR TIME.

42Appendix ECreativity Rating Form.The people listed on this page have all participated in a research study based upon your nomination. Please rate each of them according to the scale below. Indicate your rating by placing the appropriate number (1 to 11) next to each person’s name. Your ratings will be kept strictly confidential. Thank you.

EXTREMELYUNTALENTED 1 2 3 4 5 6AND/ORUNCREATIVE

EXTREMELY 7 8 9 10 11 TALENTED

AND/ORCREATIVE

43Appendix FMean Scores for Each Dependent Variable bv 1Type, System, andGroup.

Rank N Rav N-D Gen B-W MAD O-SArtist 19 10.47 48 . 05 14.68 34 . 05 21.47 65.42Neurotic 15 10 .33 49 . 60 14 .93 33 .87 21. 00 6 6.53Average 16 10 .19 41. 19 13 . 56 31.25 24.31 67 .44HarveySystem 1 23 9.96 + 45.22 15. 17 29 . 13 23 . 13 66 .91System 2 6 10 .27 49 . 60 14 .93 33 .87 21. 00 66. 53System 3 2 11 . 50 56 . 00 16 . 00 30 . 50 17 . 50 63 . 00System 4 19 10,84+ 46.74 13 . 16 34.21 18 . 79 65 .95CreativitvGroup 1 14 9 .71 49.79 15. 00 32 . 07 20.29 70.50+Group 2 12 10.67 35.33* 12.75 36.75 26.83 66.92Group 3 13 10.69 51.38 16.23 34.31 19 .23 65 .23Group 4 11 10 .36 47 .91 13.27 29.00 23 . 27 62 .00 +

Rav=Raven Advanced Progressive Matrices N-D=Nelson-Denny Vocabulary Test Gen=Generation Anagram Task B-W=Barron-Welsh Art ScaleMAD=Manifest Anxiety and Defensiveness Scale 0-S=0ne-Syllable Words Task* indicates significant difference from other groups at the * 05 level.+ indicates significant difference from each other at the .05 level.

44Appendix GMean Competitive Choices for Rank's Personality Types byCondition and Trials in the Prisoner's Dilemma Game.

Trial 1 Trial 2 Trial 3Tvpes N Mean SD Mean SD Mean SDCooperative ConditionArtist 6 8 . 50 2 .81 10. 50 3.78 8.83 4 . 12Neurotic 5 8. 00 2 . 00 9.00 2 . 55 11. 00 3 . 74Average 6 7 . 67 3 .33 10. 00 2 .90 10.67 3 .88Competitive ConditionArtist 6 9 . 33 1. 36 10. 50 1. 64 9 . 67 2 . 16Neurotic 5 8 . 20 2 . 17 7 .40 1.95 9 . 00 2 . 55Average 5 7 . 20 1. 48 10 . 00 1. 00 8 . 20 3 .2750/50 ConditionArtist 6 8 .43 1. 90 9.43 2 . 70 8 . 71 3 . 40Neurotic 5 7.20 1. 64 11.20 3 .19 11. 80 4 . 09Average 5 9 . 00 1.87 10. 80 2 .78 11. 60 4 .28Entire Sample

50 8 .20 2 .11 9.88 2 . 64 9 .88 3.49

45Appendix HMean Competitive Choices for Harvey's Conceptual Systems byCondition and Trials in the Prisoner's Dilemma Game.

System NTrial 1

Mean SDTrial 2

Cooperative Condition System 1 10 7.80 2.82System 4 3 9.00 2.65Competitive Condition System 1 8 7.75 1.58System 4 7 8.71 2.0650/50 Condition System 1 5 7.60 1.82System 4 9 8.67 1.73Entire Sample

50 8.20 2.11

Mean SD

10.60 3.448.67 .58

9.00 2.2710.00 1.83

10.00 2.00 9.89 3.10

9 . 80 2 . 64

Trial 3 Mean SD

11.00 3.437.33 3.79

7.88 2.6410.57 1.72

11.80 3.969.67 3.54

9.88 3.49

46Appendix IMean Competitive Choices for Groups (Based on Creativity Rating) by Condition and Trials in the Prisoner's Dilemma Game.

Trial 1 Trial 2 Trial. 3Type N Mean SD Mean SD Mean SDCooperative ConditionGroup 1 8 7 . 25 3 . 01 9.38 2 . 77 10 .25 3 . 50Group 2 2 9 . 50 . 71 11. 00 4 .24 12 . 50 4 . 95Group 3 4 9. 00 2 . 00 11. 25 3 . 59 11. 50 2 . 89Group 4 3 8 . 00 3 .46 8.67 3 . 06 6 . 33 4 . 16Competit ive ConditionGroup 1 4 7 . 75 1.89 8.75 3 . 69 7 .75 4 . 27Group 2 5 8 .40 2 . 30 9.40 1. 34 8.60 1.82Group 3 3 8.33 1. 16 9.67 1. 53 9 . 67 2 . 08Group 4 4 8 . 75 2 . 06 9.75 1. 50 10. 25 1. 5050/50 ConditionGroup 1 2 6. 50 . 71 7 . 50 .71 7.50 .71Group 2 5 8 . 80 1. 92 11. 60 4 . 04 10. 40 4 . 72Group 3 6 8 . 17 2 . 14 10 . 50 1.76 11.33 4 . 68Group 4 4 8 . 50 1.73 10 . 00 2 . 58 10.75 2 . 87Entire: Sample

50 8.20 2 . 11 9.88 2 . 64 9.88 3 . 49

47Appendix JAnalysis of Variance Table for the Performance of Rank'sTypes in the Prisoner's Dilemma Game.

Source of Variation ss DF MS F PBetween Subiects.PERSONALITY TYPE (A) 1. 56 2 .78 . 05 .955CONDITION (B) 22 . 61 2 11. 31 . 67 . 515A X B 50. 22 4 12.56 .75 . 565ERROR BETWEEN 687.93 41 16. 78Within Subiects.TRIAL BLOCKS (C) 99 . 22 2 49 . 61 13 . 05 . 000A X C 39.52 4 9.88 2 . 60 . 042B X C 14 .93 4 3 . 73 . 98 . 422A X B X C 35 . 13 8 4 . 39 1. 16 . 336ERROR WITHIN 311.60 82 3 . 80

48Appendix KAnalysis of Variance Table for the Performance of Harvey'sConceptual Systems in the Prisoner's Dilemma Game.

Source of Variation SS DF MS F PBetween Subiects.SYSTEM TYPE (A) .28 1 . 28 . 02 .892CONDITION (B) 8 .80 2 4 .40 .30 . 745A X B 42 . 98 2 21.49 1.45 .248ERROR BETWEEN 534.15 36 14 .84Within Subiects.TRIAL BLOCKS (C) 51. 37 2 24 . 75 7 . 08 . 002A X C 22 .16 2 10. 31 2 .95 . 059B X C 11.97 4 3.17 .91 .464A X B X C 41.25 4 9.80 2.80 . 032ERROR WITHIN 251.59 72 3 . 49

49Appendix LAnalysis of Variance Table for the Performance of GroupsBased on Creativity Ratings in the Prisoner's Dilemma Game.

Source of Variation SS DF MS F PBetween Subiects.GROUP (A) 79.74 3 26 . 58 1. 64 . 197CONDITION (B) 8 . 69 2 4 . 35 . 27 . 766A X B 66. 69 6 11. 11 . 69 . 663ERROR BETWEEN 616.35 38 16 . 22Within Subiects.TRIAL BLOCKS (C) 65 . 78 2 32.89 7 .25 . 001A X C 8 . 16 6 1.36 .30 . 935B X C 6 . 55 4 1. 64 . 36 . 836A X B X C 33 . 68 12 2 . 81 . 62 . 820ERROR WITHIN 344.83 76 4 . 54